Letter from the CTO (Nov 2020) - Can work automation solve the efficiency crisis?

[#1]

Hi! My name is David Burggraaf and I’m the Chief Technology Officer at Workfront. I’m excited to be sharing some thoughts with you this month on the topic of work automation. Throughout my career, one of the things I have observed is that those that do well, automate. Whether it is engineers or people in finance, marketing, sales, HR, IT, or elsewhere, one of the common traits of the successful is that they automate. So, when a brother of mine asked me how they could be more successful in their career, I said “automate!”

Before we dive in, it’s important to set the stage. Nearly every organization struggles with efficiencies on some level. Everyone is trying to get products to market faster while still being innovative. It’s not a surprise when we see that over 40% of employees are spending at least a quarter of their week doing repetitive tasks. A recent CMO Council study noted that $1 trillion is lost by companies every year due to mismanaged tasks, resulting in wasted productivity. And half of the average knowledge worker’s time is spent in business reviews, project update meetings, and other tasks focused on the processes related to work—rather than the work itself. Something has to give. 

Historically, we’ve seen several ways organizations try to solve this problem. One obvious answer is to simply add more resources, but we all know budgets aren’t quite as simple as that. You might take the path of doing nothing, and just asking people to “work harder.” That might work well in the short term, but it’s not sustainable long term. There has been an emergence of new technologies that try to solve the pain of inefficiencies through automation. We see people automating with a traditional iPaaS solution, or integration platforms as a service. We’ve also seen the emergence of robotic process automation, which is meant to mimic human behaviors and interact with systems in that particular way.

In addition to all of these solutions (or lack thereof), we believe that there’s an entirely new category of solution known as knowledge work automation, and that it can solve a great deal, if not all, of this efficiency problem.

Knowledge Work Automation

First off, what is “knowledge work”? This is the kind of work that can be differentiated from other forms of work by its emphasis on “non-routine” problem solving. It requires a combination of convergent, divergent, and creative thinking. 

The concept of automating knowledge work encompasses a wide variety of activities, from very basic, such as automating system tasks, to revolutionary, such as automating decision-making. 

To think through all the variables that are involved in automating knowledge work, we have created a maturity model. This model includes several stages of automation that describe the task to be done, the efforts required to implement, and the benefits to your organization.  

• Automating Workfront. The most common type of automation related to Workfront is the automation of Workfront itself! By automating system processes, the administration of the platform is done without manual effort and allows people to concentrate on their work rather than the system. A great example of this type of automation is when customers automatically provision or deprovision Workfront users, and ensure those users have the correct settings to be successful in using Workfront. Or it may include a process to ensure your Workfront data is well organized by automatically attaching the right forms and approval steps to projects and tasks. These automations are a great place for teams to start because these processes are typically well-defined and you can easily quantify the benefits of removing the manual effort from someone’s job. 

• Automating Workflows. The next level of maturity involves the process of automating the hand offs of work between both people and systems. It’s the “work between the work” if you will. An example of this type of automation would be something many Workfront customers do today, which is converting a request for work into actionable projects that can then be executed by other people. There is quite a bit of variation of maturity here, from simple workflow automation to much more complex logic. You might start by converting all requests for work into projects, and then adjust to only convert the request if it includes certain thresholds, or to convert it and then assign it to a specific team based on the data in the request. This level of maturity also includes an entire category of automation that transfers data from one system, typically used by one set of stakeholders, to an entirely different system, utilized by a completely different set of stakeholders. These processes are meant to integrate your entire lifecycle of work in order to increase data quality, decrease coordination time, and limit the introduction of errors. 

• Automating Work. This next level of automation maturity is where things start to get exciting. At this level of maturity we’re thinking beyond automating systems or the coordination of work, and we’re thinking specifically about how we can automate the work itself. This is about automating the “value-added” things that a person was originally hired to do. Imagine a scenario where a creative design professional doesn’t just receive a task to create a new InDesign layout, but instead they receive an InDesign file that is pre-populated with the right file template including sizing, ready artwork, and slugs. Or imagine a software engineer who is assigned a bug to work on, and the correct code repository is automatically pulled into their local environment so they can begin the coding enhancement or fix immediately. These types of automations are on the frontier of innovation within knowledge work. They are less prominent today, as the opportunities can be harder to identify and implement, but the impact could be significant. 

Much of the types of automation we have described up to this point are heavily dependent upon a strong process orientation. It’s absolutely critical that you understand your current processes and ensure they are the RIGHT processes before automating them. The next few stages of maturity will build on this foundation of process-driven automation, and data will become even more critical. 

• Automating Insights. The level of maturity focuses on teaching a system how to make accurate predictions when fed data. How do we take something that would normally require a lot of time or human intervention to determine, and automate it based upon patterns in data. An example of this would be automatically identifying which teams are overloaded even though work is often hard to quantify, or automatically determining what caused a project to go off the rails, or even learning to automatically scan completely new marketing materials to determine if they meet legal and branding expectations. These are things we’re experimenting with at Workfront, and that we’re planning more deeply for the future. 

• Automating Decisioning. The last level of knowledge work automation maturity is something that is often talked about, but rarely done to the full extent by organizations, and that’s artificial intelligence. Artificial intelligence is the act of automating the crown jewel of knowledge work—automating not just insights, but the act of making decisions with data that backs up those decisions. Early innovation in this area includes potentially automating changing work assignments to ensure the best people are always appropriately utilized on the most important work, and allowing barriers to be overcome without human intervention. It’s about not just identifying potentially problematic marketing material, but actually automating the decision on which material would be most effective to accomplish specific results. 

Some of the elements in this maturity model are fully within reach today, and some are audacious. This framework was created to get us thinking—am I automating the doing or am I automating the thinking? In order to automate the thinking, we have to get really good at the doing, and collect the right data in the process. 

We have a number of customers who are leveraging Workfront Fusion and Workfront itself to automate much of their process-driven work today. If you’re interested in learning more about how other customers are automating workflows and integrating systems, I encourage you to join the Fusion 2.0 Group on the Workfront One community. It’s a group of Workfront professional service partners, Workfront team members, and customers who are engaged in learning and mastering the new Workfront Fusion technology. 

I hope this article planted some seeds of inspiration. Thanks, as always, for being part of the Workfront family and for helping us grow and be better every day. If there’s anything we can do to help, or to make your day-to-day life easier, please don’t hesitate to ask.